Stephen Jones — Robotics & Mechatronics

Georgia Tech B.S. in Mechanical Engineering ('26)
Georgia Tech M.S. in Computer Science — Computational Perception & Robotics (Enrolled)
Division I Varsity Swim Team Captain (2024–2025)

Live Portfolio: sjones-code.github.io
LinkedIn: linkedin.com/in/sjones396
Email: stephen04317@gmail.com

---

Technical Core Competencies

• Mechatronics & Firmware: ESP32, Embedded C/C++, Arduino, PWM regulation, logic-level MOSFET switching, Shape Memory Alloys (Nitinol), I2C / SPI / UART, rapid soldering & harness wiring.
• Mechanical Design & Prototyping: SolidWorks, AutoCAD, Bluebeam Revu, FMEA risk analysis, QFD / House of Quality, pneumatic blowdown systems, 3D printing & additive fabrication.
• Perception & Software: Python, C++, OpenCV, TensorFlow, Scikit-learn, MATLAB, Linux / Bash (In-Training), ROS 2 (In-Training).

---

Featured Engineering Builds

1. Wireless Shape Memory Alloy (Nitinol) Robotic Claw (Capstone)

• Engineered a noiseless, high power-to-weight robotic gripper utilizing Joule-heated Nitinol wire instead of electric gearmotors.
• Programmed an ESP32 microcontroller hosting a standalone local Wi-Fi web server for smartphone control with sub-50ms latency.
• Designed PWM-driven MOSFET circuits to dynamically regulate current and prevent thermal alloy fatigue.

2. Autonomous Competition Mobile Robot (ME 2110 — 1st Place Champion)

• Designed, machined, wired, and programmed an autonomous mobile robot from scratch under strict dimensional and timing rules.
• Placed 1st overall out of 5 teams in Georgia Tech's final arena showdown.
• Handled 30+ weekly hours of SolidWorks modeling, rapid 3D printing, laser cutting, and deterministic C++ control coding.

3. CO2 Blowdown Wildfire Suppression Drone Payload (Better Place Drones)

• Designed a regulated pneumatic blowdown payload system using CO2 canisters to pressurize a water reservoir for instant fire suppression.
• Coordinated with 5 subteams (airframe, telemetry, avionics, payload, recovery) to deliver physical flight-ready test modules.
• Advanced to the Top 30 teams internationally out of 300 entries in the global XPRIZE Wildfire challenge.

4. Deep Learning Concrete Defect Classifier (ME 4710)

• Fine-tuned a pre-trained VGG16 CNN as a deep feature extractor, converting 2,000 structural concrete images into 512-dimensional feature tensors.
• Performed Principal Component Analysis (PCA) to capture 95% of dataset variance within 104 components, reducing feature dimensionality by 80%.
• Achieved 99.67% test accuracy on 600 validation images with 1.00 precision and recall across classes.

---

Repository Structure

sjones-code.github.io/
├── assets/
│   ├── Better Place Drones/    # System CAD, section views, and flight airframe photos
│   ├── Brandon Adams/          # Leadership and teammate award photos
│   ├── Capstone/               # Prototyping progression, ratchet sub-assemblies, and actuation video
│   ├── ME 2110/                # Competition robot CAD, arena photos, and final presentation PDF
│   ├── Vision Perception/      # Confusion matrix, accuracy plots, and crack morphology clusters
│   ├── GTVertical_RGB_0.png    # Georgia Tech emblem
│   ├── headshot_square.jpeg    # Profile avatar
│   └── Stephen Fall 26' Resume.pdf # Official engineering resume
├── index.html                  # Interactive, accessible single-page portfolio
├── style.css                   # Responsive layout system, blueprint theme, and dark mode tokens
└── README.md                   # Engineering overview and documentation